Re: wind instrument +- filtering of broadband noise

[John Denker <jsd@MONMOUTH.COM>]

At 03:08 PM 12/9/00 -0500, Tom McDonald wrote:

> I do the probably common demonstration in which a large lighted laboratory
> gas burner, e.g. Fisher burner, is placed on the floor and a long cardboard
> carpet roll  tube is lowered over it until a very loud low pitched sound is
> produced.

OK.

>  Is it true that even though this clearly IS a hissing source,
> there is more going on than "a resonant filter applied to a broadband noise
> source"

Yes, there is a lot more going on.  You could redesign the burner to get
rid of the hissing, but the loud sound would persist.

For those who aren't super-familiar with a fisher burner:  at the bottom it
has a gas orifice and an air intake, constituting the carburetor.  Above
that there is a rather long fat tube in which the fuel-air mixture does
_not_ burn.  Then there is a grid, above which the flame sits.

In the carpet-tube demonstration, the essential physics is as follows:  The
resonant air-motion in the tube assists the flow of fuel/air mixture out of
the non-burning region into the burning region during part of the cycle,
and opposes the flow during the opposite part of the cycle.  This means
that although a steady supply of gas is provided, it is consumed
unsteadily.  The power in a flame is very very large, so even a modest
coupling results in quite a lot of acoustical power.

> Is this "mode locking"?

The description given above does not permit a firm conclusion as to whether
mode-locking is or is not occurring.

  a) Under some conditions it should be possible for two incommensurable
modes to persist, both modulating the burning-rate.

  b) Under other conditions it would not be surprising if mode-locking did
occur, that is, that one mode would do something nonlinear to the energy
source so that there would be nothing left to feed other modes.